Voice in the business process

Thomas Howe gave a fascinating session on how it is possible to use Web Services to enhance customer service. He demonstrated a triage on-call service for doctors that utilized Amazon's Mechanical Turk Artificial artificial intelligence service to triage patient calls. The demonstration he showed utilized VoiceXML, Ruby on Rails, Service Oriented Architecture and some of the Amazon Services and was developed in 40 hours by one person. For anyone not familiar with Mechanical Turk, it is artificial, artificial intelligence. Amazon have mobilized human expertise and made it available as a web service. Examples are services like CastingWords. CastingWords will take an uploaded voice file, say from an interview or training session and break it in to bite size chunks and send this out to Mechanical Turk where people will receive the audio segment and transcribe it. These are returned to CastingWords where they stitch together the pices of the transcription and send it back to the customer. The customer pays CastingWords for the service and CastingWords pays a set fee for each audio segment sent to Mechanical Turk. Amazon handles the payment for services rendered to all their "Turks". The IVR example used this capability with a team of qualified nurses registered on the Mechanical Turk. The critical lessons from this demonstration are:

1. The ease and speed of integration
2. The scalability of the solution
3. The cost effectiveness of the solution
4. The ease of integration in to business processes with no customer education required.